Abstract
The image quality evaluation method, based on the convolutional neural network (CNN), achieved good evaluation performance. However, this method can easily lead the visual quality of image sub-blocks to change with the spatial position after the image is processed by various distortions. Consequently, the visual quality of the entire image is difficult to reflect objectively. On this basis, this study combines wavelet transform and CNN method to propose an image quality evaluation method based on wavelet CNN. The low-frequency, horizontal, vertical, and diagonal sub-band images decomposed by wavelet transform are selected as the inputs of convolution neural network. The feature information in multiple directions is extracted by convolution neural network. Then, the information entropy of each sub-band image is calculated and used as the weight of each sub-band image quality. Finally, the quality evaluation values of four sub-band images are weighted and fused to obtain the visual quality values of the entire image. Experimental results show that the proposed method gains advantage from the global and local information of the image, thereby further improving its effectiveness and generalization.
Highlights
Image quality evaluation has a wide range of applications in image compression, image restoration, and video processing
Before using the convolutional neural network (CNN) to predict the quality of the sub-band image, the image is normalized for local contrast, i.e., removing redundant features that are weakly related to image quality
The evaluation methods include the structural similarity methods proposed in [4], the DIIVINE evaluation method based on natural scene statistics in [10], the BRISQUE evaluation method based on spatial domain in [11], the BLIINDS-II method based on DCT in [12], the evaluation method based on CNN in [15], and the BIECON evaluation method in [16]
Summary
Image quality evaluation has a wide range of applications in image compression, image restoration, and video processing. Bare et al [17] used 32 × 32 image blocks as inputs to the CNN similar to the network framework proposed in [16] and adopted the full reference quality evaluation algorithm [18] to calculate the quality score of the image block as the output of the CNN. The existing image quality evaluation method based on CNN uses the average value of the image sub-block to represent the quality evaluation value of the entire image This method can detect lowand high-quality image regions and achieve good image quality evaluation results. We adopt the information entropy as the weight of quality prediction of sub-band image, and demonstrate that the distribution of information entropy is close to the image region of human visual perception Using this strategy, the subjective and objective consistencies of the image quality evaluation can be further improved
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
